I am a biomechanics researcher, and wouldn't say that anything in that article is exactly news. This isn't a criticism of Hutch, he's just brought attention to it outside of academia which is a great thing.
While the specific study discussed is novel, it's been established for quite a while that someone's specific running or walking biomechanics already represent an energetic optimization of the current variables at play ("requested" speed, incline, ground/shoe stiffness, active ROM, muscular strength/endurance etc.). You can see this from studies where these variables are controlled or changed - we respond very quickly to find the lowest energy style possible within the free variables available to us. This is an evolutionary feature of humans - we are a species designed to walk or run long distances, and doing so inefficiently has consequences. By imposing artificial constraints (ie. certain aesthetic styles) on someone you're likely moving them further from the optimal.
I wouldn't say that it's totally incorrect to try to change someone's form. While one's nervous system adopts the most optimal form, this is based on "current conditions," which aren't necessarily the most optimal for that person. So for example, someone with weak abdominal muscles would run in the way that is most efficient given this weakness, but it is possible that their absolute efficiency would be improved by "unlocking" different form possibilities through a weights/core program. So working on stuff like ROM, neuromuscular activation, power, strength etc. through non-running ancillary training might help improve absolute running economy for an individual.
This is a subtlety that is important but often missed. A coach or a biomechanist or an algorithm cannot identify/prescribe what specify global parameters make a runner efficient because a runner's style is a consequence of complicated interactions in an individual's physiological set-up. But, this individual's set-up can be improved to some extent to allow them to run in a way that is more efficient.